The gap effect: discontinuities of preferences around dead.

BACKGROUND The assessment of health states considered to be worse than dead is a controversial issue. OBJECTIVE To investigate how health states are valued when they are close to dead. Differences between adjacent states are compared with the differences between the first positive/first negative state with death. METHODS A secondary analysis of the EuroQol EQ-5D data of the measurement of valuation and health (MVH) study was made. Visual analog scale (VAS) and time trade-off (TTO) preferences for 43 health states were obtained. Various subsets of 13 states were valued by 3395 respondents. States were rank ordered by their VAS and TTO values. Differences between adjacent states were calculated for the VAS and the positive and negative TTO values. RESULTS Complete data were obtained in 2997 respondents. The differences between the ordered VAS values were equally large. In contrast, significant gaps around dead were found for the positive as well as the negative TTO values. DISCUSSION These results are interpreted in light of a descriptive QALY model. This model was expanded to include utilities worse than dead. The VAS task does not pick up that bad states become intolerable, i.e. worse than dead, when they last too long, but the TTO task does. The current QALY model seems to lack descriptive validity for states valued worse than dead and for states with a maimal endurable time.